Business forecasting · Practical planning · Senior advisory
Forecasting Consultant & Advisor. Make the next decision with a clearer view ahead.
Connect demand, revenue and capacity before the pressure reaches your team.
I help business leaders build forecasts they can understand, challenge and use. We start with the decision you need to make, examine the evidence and assumptions, then connect the outlook to a practical plan for the people doing the work.
The purpose of forecasting consulting
A useful outlook changes what you do next.
The forecast should answer a business question precisely enough to support a decision.
A forecasting consultant helps an organization estimate future activity over a defined period, understand the uncertainty around that estimate and use it to plan resources and actions.
The work can involve demand, sales, revenue, workloads, inventory or service capacity. It also involves definitions, data quality, decision timing and the people responsible for acting on the result. A mathematically sound projection still needs a place in the operating rhythm of the business.
My role is to connect those parts. I work with leadership, finance, sales, operations and analytical specialists to define what needs to be forecast, what evidence is available and what should happen when the outlook changes.
Sometimes that means rebuilding an unreliable planning process. Sometimes it means creating a first practical model, improving an existing forecast or helping decision makers challenge assumptions without turning the meeting into a contest over whose spreadsheet gets the final word.
We can begin with the problem you are trying to solve. You do not need to arrive knowing which model, platform or forecasting method to request.
When forecasting needs attention
The symptoms usually appear in the business first.
You may already have extensive reporting and capable people. The missing piece is often a shared, credible view of what is coming and enough time to respond.
Revenue surprises
The pipeline looks healthy. The period ends differently.
Opportunities slip, close rates vary, renewals are assumed and bookings are confused with revenue. The forecast needs explicit timing, definitions and evidence behind the assumptions.
Capacity pressure
Growth reaches the calendar before it reaches the plan.
Sales commitments create work that available people, equipment or locations cannot absorb. Demand needs to be translated into workload, skill requirements and realistic delivery timing.
Planning disagreement
Every department has a reasonable explanation.
Finance, sales and operations may be measuring different things over different periods. A useful process makes the differences visible and reconciles what can be reconciled.
Inventory imbalance
The total looks acceptable. The mix is wrong.
Aggregate sales can conceal shortages in one product and excess stock in another. The forecast needs the right level of detail for purchasing, replenishment and delivery decisions.
Weak assumptions
A target has quietly become the expected outcome.
Ambition belongs in the plan. The outlook also needs evidence about demand, conversion, timing and resources. Otherwise, the gap remains hidden until it becomes expensive.
No response
The forecast changes. Nobody knows what follows.
The organization needs decision thresholds, responsible people and a review cadence. An updated number should make the next conversation more useful.
Start with clear definitions
The target, forecast, scenario and plan do different jobs.
A leadership team can reasonably want an outcome that the current evidence does not support. The productive response is to make that gap explicit and decide what would have to change.
Suppose the target is $500,000 in new bookings and the current base scenario supports $360,000. The resulting $140,000 gap creates a useful planning question. Which combination of qualified demand, conversion, project value and timing could close it? What resources would that require, and how will the team know whether the plan is working?
Changing the forecast to $500,000 because the target matters removes information the business needs. The forecast should be allowed to tell leadership something inconvenient.
Target
The result you want to achieve, with a defined measure and deadline.
Forecast
An estimate of future outcomes based on available evidence, a method and stated assumptions.
Scenario
A conditional view of what could happen if particular assumptions or events occur.
Plan
The actions, resources and responsibilities intended to influence the outcome or prepare for it.
Keeping these distinctions clear also makes accountability fairer. A team should be able to update an outlook when evidence changes without being accused of abandoning the goal. It should then explain the implications and the response.
Business forecasting consulting services
Build the outlook around the decision it must support.
We can focus on one high-value problem or connect several forecasting needs. Scope follows the operating decision, the available data and the cost of being wrong.
Demand forecasting
What activity should the business prepare for?
Estimate demand by the product, service, location, channel or period that matters. Examine seasonality, promotions, lost demand, changes in customer mix and the constraints that affect what your historical sales actually represent. The result should inform a specific resource or purchasing decision.
Sales & revenue forecasting
What is likely to close, when, and with what consequence?
Review pipeline stages, time in stage, expected start dates, deal size, renewals and conversion assumptions. Keep new bookings, delivery, recognized revenue and cash receipts distinct, with finance defining the appropriate treatment. A single total rarely explains the timing the business needs.
Staffing & capacity forecasting
Can the organization deliver the work it expects?
Translate anticipated activity into productive hours, required skills, equipment use or other meaningful capacity measures. Account for the work already committed, availability, ramp-up time and service standards. Review scenarios before turning a projected increase into a hiring or scheduling commitment.
Inventory & operational planning
Where does timing create avoidable pressure?
Connect demand assumptions with replenishment lead times, product mix, supplier variability and operational constraints. A forecast is one input to inventory and production decisions; policy, cost and service requirements also need to be explicit.
Resource & expense scenarios
What changes when the operating assumptions change?
Connect activity levels to relevant resource needs and costs. Work with your finance team on definitions, payment timing and planning implications. The aim is to show which assumptions materially affect a business decision and where further specialist analysis is appropriate.
Forecasting process design
How does the outlook stay useful?
Define source systems, responsibilities, refresh timing, version history, review meetings and escalation rules. Establish how departmental forecasts connect and how assumptions are challenged. The process should be manageable by the people expected to maintain it.
Start where the value is clearest
A bounded question such as “How much delivery capacity will we need over the next eight weeks?” gives us a practical starting point. Once the approach is useful and maintainable, we can decide whether other forecasting needs justify expansion.
A worked business example
When stronger sales create a delivery decision.
Consider a hypothetical project-based service company. Each won project has an assumed booking value of $12,000 and requires 40 productive delivery hours. Available capacity is 1,200 productive hours in the planning period.
For this simplified example, there is no opening backlog and all won work is scheduled within the same period. Actual planning would also account for project mix, timing, cancellations, rework and existing commitments. These scenarios are assumptions, not probabilities or client results.
On a small screen, scroll the table horizontally to compare every scenario.
| Planning input or output | Lower demand | Base scenario | Higher demand |
|---|---|---|---|
| Qualified opportunities | 80 | 100 | 120 |
| Assumed close rate | 25% | 30% | 35% |
| Implied new projects | 20 | 30 | 42 |
| Implied booking value | $240,000 | $360,000 | $504,000 |
| Required productive hours | 800 | 1,200 | 1,680 |
| Available hours minus required hours | 400 available | 0 available | 480 short |
Implied projects × productive hours per project = delivery workload
The operational question becomes visible.
The higher-demand scenario produces 12 more projects than the base scenario and requires 480 additional productive hours. Leadership can now examine scheduling, work selection, start-date commitments, qualified subcontracting or other appropriate capacity options.
The base scenario also deserves attention. It uses the full assumed capacity, leaving no allowance in this simplified example for disruption or estimation error. Whether that is acceptable depends on the service model and the consequences of delay.
The next test is the quality of the assumptions.
Are the opportunities genuinely qualified? Does the close rate reflect comparable work? Is 40 hours a reasonable estimate for the expected mix? Is the available capacity measured in productive delivery hours rather than total paid hours?
Booking value is not profit, recognized revenue or cash collected. Those require their own definitions and timing. The model becomes useful when the team understands what each number means and what it leaves out.
How I approach the work
Start with the people who must make and carry out the decision.
Forecasting projects should begin with the relevant decision makers in a conversation. We need to understand the underlying goal, the current problem, the resources available and the practical consequences of the answer.
Agree on one measurable objective.
Define the decision, measure, planning horizon and date by which the answer becomes useful. For example: improve the eight-week workload outlook enough to support a weekly capacity review. Establish how we will assess improvement before choosing a model.
Assess the team and the constraints.
Review the strengths, weaknesses, opportunities and threats of the people assigned. Identify missing expertise, conflicting responsibilities, available hours and likely obstacles. Include the people who maintain the data and those who will use the forecast.
Audit the definitions and available evidence.
Trace important measures back to their sources. Establish which events are recorded, when records become available and what has changed over time. Agree how to handle missing information, unusual periods and conflicting interpretations.
Build a baseline and test the alternatives.
Start with an understandable reference forecast. Compare other approaches using an evaluation that respects time and matches the business horizon. Assess the assumptions, maintenance burden and decision value alongside numerical accuracy.
Connect the outlook to actions.
Define what changes when the forecast crosses a meaningful threshold, who reviews the signal and what evidence is needed before acting. Make space for uncertainty and appropriate human judgment.
Review outcomes and maintain the process.
Compare the forecast issued at the time with what subsequently happened. Record adjustments and their reasons. Improve the model, the assumptions and the decision process as the business learns.
Data readiness and business reality
A clean spreadsheet can still describe the wrong thing.
Before fitting a model, I want to know how the numbers were created. The difference between a request, a qualified opportunity, an accepted estimate, a scheduled job and a completed job matters. So does the difference between the date an event occurred and the date someone entered it.
Separate observed activity from the activity you could not serve.
Imagine a service business that stops accepting appointments once its calendar fills. Recorded jobs describe fulfilled activity. They do not capture every customer who wanted an appointment. Without inquiry or turn-away information, the team may mistake a capacity ceiling for a demand pattern.
The same question applies when a product is unavailable or a location is closed. We should identify where history reflects a constraint and determine what additional evidence is available. If the missing demand cannot be estimated credibly, that limitation belongs in the decision discussion.
Record changes that affect comparability.
Pricing, service territory, product mix, working hours, marketing activity and qualification rules can all change what a historical period means. A monthly total may look consistent while the underlying business has changed substantially.
I would ask the team to maintain a short event log beside the data: what changed, when, who knows the details and why it might affect the outlook. That context helps an analyst avoid treating every unusual observation as either an error to delete or a permanent new trend.
Use the information that existed at the decision date.
Suppose a sales opportunity was marked “won” after the period ended. An evaluation using its current status to predict that earlier period would give the model information the business did not have at the time. The result could look impressive while providing an unrealistic picture of performance.
Preserve snapshots or otherwise reconstruct what was knowable when each forecast was issued. Forecasting has a time dimension in both the outcome and the evidence. We need to respect both.
Make data access and maintenance workable.
Agree who can provide the necessary records, which fields are actually needed and where the working data belongs. Document definitions and transformations so the process does not depend on one person remembering an undocumented sequence of exports.
If the main problem is understanding why performance has changed, my plain-English guide to business analytics and root-cause analysis offers a useful starting point. An explanation of the past and a forecast of the future should inform each other without being treated as the same task.
Methods that earn their place
Choose the approach around the evidence and the decision.
The method needs to match the time horizon, data volume, stability, level of detail and maintenance capacity. A tool is useful when it improves the forecast or the decisions it supports enough to justify its complexity.
Start with a reference
Simple baseline forecasts
Use an appropriate reference such as the latest comparable period, a seasonal comparison or another simple rule. Hyndman and Athanasopoulos explain common benchmarks. A proposed improvement should be assessed against a baseline worth comparing with.
Explain the mechanism
Driver-based planning
Connect the quantities that generate an outcome, such as qualified opportunities, conversion, project value and delivery hours. The worked example above is a simple illustration. Each driver needs a definition, evidence and an appropriate review process.
Learn from sequence
Time-series approaches
Where suitable history exists, evaluate methods that represent patterns over time. Assess performance at the actual planning horizon and against the reference forecast. Document what happens when the pattern or business conditions change.
Use additional information
Models with explanatory variables
Consider relevant information available when the forecast must be made. Time-series regression offers one framework. Future predictor values may themselves be uncertain, and an observed relationship does not automatically establish cause and effect.
Explore alternatives
Scenarios and informed judgment
Use structured assumptions when evaluating a new offer, a major operational change or another situation with limited comparable history. Judgmental forecasting can be part of the process; assumptions and revisions still need to be documented.
Keep totals coherent
Connected operational forecasts
When forecasts exist by branch, product or region, examine how they relate to company totals. Hierarchical and grouped forecasting addresses this consistency problem. The right level of detail depends on the decisions each part of the organization makes.
Trust through evaluation
Test the forecast the way the business will use it.
An impressive fit to past data is not enough. We need to know how an approach performs when asked to estimate outcomes it has not yet seen, using information that would actually have been available.
Time-series cross-validation tests forecasts across successive historical decision points. The training information precedes the outcomes being evaluated. The forecast horizon should match the business need: an approach that works one week ahead may behave differently over the eight weeks needed for a staffing decision.
Measure the size and direction of errors.
Typical error size helps describe how far the forecast misses. Bias helps identify whether it consistently runs high or low. Segment-level review can show whether a reasonable company total conceals persistent errors in a branch, service or product.
Metric choice matters. Percentage errors can be unstable or undefined when actual values are near zero or zero. This guide to forecast accuracy measures explains why no single percentage should become a universal grade. Choose measures that fit the data and interpret them alongside business consequences.
Evaluate the decisions as well as the numbers.
Underestimating demand may create missed service commitments; overestimating it may leave resources underused. The costs are not necessarily equal. Leadership should understand the tradeoff when choosing how to act on the outlook.
For a capacity decision, I would want to review whether the process helped identify pressure in time to respond, whether the response was practical and whether the evidence improved over repeated cycles. A smaller numerical error is useful when it changes something the business values.
Keep the forecast that was actually issued.
Archive the original version, the date, the assumptions and any later adjustments. Comparing actual outcomes with a retrospectively edited forecast can make the process look better while hiding the lesson.
Review the baseline, the model output and the final human-adjusted forecast where relevant. If an adjustment helped, understand why. If it did not, use that evidence to improve the next review rather than making the number a matter of personal credibility.
Uncertainty that people can use
A range needs an explanation.
A statistical prediction interval and a set of business scenarios answer different questions. A prediction interval describes uncertainty under a model and its assumptions. A scenario describes an outcome conditional on selected assumptions. Labeling an optimistic and pessimistic case does not automatically make them a statistically calibrated range.
Prediction intervals help communicate the spread of possible outcomes and can be evaluated against subsequent observations. They do not remove the possibility that the model omits a disruption or that the business changes in a way the historical data does not represent.
For practical planning, I want leadership to understand what is covered, what remains outside the analysis and which uncertainty matters for the decision. If a supplier delay changes the result materially, that deserves a clear scenario and an action discussion.
The useful question is what we would do differently if the outlook moved.
A response might involve reviewing commitments, gathering more evidence, preserving flexibility or preparing an option that can be activated later. The forecast should support an appropriate decision under uncertainty, with a clear explanation of the assumptions behind it.
The business context changes the forecast
Different industries need different planning questions.
These are illustrative situations, not client case studies or promises of results. Each shows how I would connect a forecasting question to the people and decisions around it.
A professional services firm is selling work faster than it can start it.
The pipeline contains several attractive projects, but they require different specialties and have uncertain start dates. A total revenue forecast gives the partners an incomplete view of delivery pressure.
I would separate committed work from potential work, document timing assumptions and translate likely projects into demand for the relevant skills. The review should show where a delay in one engagement or an earlier start in another changes the capacity picture.
The decision might concern staffing, project sequencing or commitments made during sales discussions. That operating view connects naturally with professional services growth consulting, where the quality of the work and the capacity to deliver it are part of the commercial offer.
A manufacturer has acceptable total demand and the wrong production mix.
Company-level sales are close to plan, but one product family needs scarce machine time while another leaves available capacity unused. Purchasing and scheduling need more detail than the total provides.
I would examine the planning level, lead times, product relationships and constraints. A useful forecast would help the team see the workload implications of the expected mix, while recognizing the limits of what history can support at very fine levels of detail.
The next decision could involve scheduling, procurement or a discussion with customers about timing. The wider commercial context is covered in my manufacturing, industrial and supply-chain consulting.
A service business is getting more inquiries without more profitable work.
The owner sees rising lead volume and expects revenue to rise with it. Some inquiries are outside the service area, some concern unsuitable projects and some never become estimates. Forecasting from raw lead count hides those differences.
I would define the stages that matter, distinguish project types and examine the timing between inquiry, qualification, estimate, acceptance and delivery. The model should use evidence that reflects the work the company actually wants and can perform.
Leadership can then assess whether the next useful action concerns audience selection, follow-up, estimating capacity or the forecast itself. Increasing activity in the wrong part of the process may add workload without improving the outcome.
A museum or attraction needs an outlook that respects its calendar.
A monthly attendance total may conceal school groups, special events, member visits and individual visitors with different operating needs. Changes in opening hours or an unusually large event can also complicate historical comparisons.
I would start with the decisions staff need to make: visitor service coverage, program capacity, purchasing or promotion timing. Then I would determine which audience segments and time periods are useful and what information is available before each decision.
The broader museum and cultural attraction strategy should help interpret the outlook. An anticipated audience needs an experience the institution can deliver well.
What the engagement can produce
A usable forecasting process, with the reasoning included.
We agree deliverables around the scope. The goal is to leave your team with work it can understand, operate and improve.
Direction
A decision and forecasting brief
The objective, measure, horizon, frequency, responsible people and criteria for evaluating whether the work is useful.
Evidence
A data and assumption review
Definitions, source limitations, missing information, important business changes and the assumptions that need regular attention.
A working approach
A baseline and model specification
An understandable forecasting approach, with the appropriate working model or a clear implementation specification for your analytical team.
Confidence
A validation and scenario review
What was tested, how it performed, which limitations remain and how alternative assumptions affect the planning discussion.
Action
A forecast-to-decision plan
The signals to review, available responses, approval responsibilities and the evidence needed before committing resources.
Continuity
Documentation and a review rhythm
Refresh instructions, version history, responsibilities and a practical method for learning from subsequent results.
The meeting after the model
Make the forecast part of an honest operating conversation.
A forecast review should have a defined purpose. The team needs to understand what changed, why it changed, whether the change is meaningful and what response is appropriate.
I would structure the conversation around the current outlook, important differences from the previous version, the assumptions driving those differences and any decision approaching its deadline. The people with relevant context should be able to contribute before the number becomes an executive summary.
Sales may know that a buyer’s approval process has slowed. Operations may know that a resource will be unavailable. Finance may know that the timing in a report does not match the timing needed for planning. These are useful contributions when the process records them clearly and evaluates their effect.
Every adjustment should have a reason, an owner and a date. Where feasible, retain the unadjusted forecast as well so the team can learn whether its judgment improved the result. Over time, that creates a more useful discussion than treating every change as either a triumph or a failure.
The cadence should match how quickly the information changes and how much lead time a response needs. A weekly review might be appropriate for one operational question and wasteful for another. We design the rhythm around the decision.
Working together
Senior help at the level your team needs.
I work directly with business owners, founders and leadership teams. We can begin with a defined problem and choose a sensible engagement around it.
Focused project
Diagnose or rebuild one forecasting process.
Clarify the problem, assess readiness, develop the approach and create a practical handoff. Useful when a particular decision or recurring issue has become difficult to manage.
Retained advisory
Improve the outlook over repeated planning cycles.
Support reviews, challenge assumptions, examine results and help leadership adapt the process as the business changes.
Embedded support
Connect leadership with the team doing the work.
Provide senior direction and coordination alongside finance, operations, sales and analytical specialists. Remote collaboration and on-site work can be considered around the need.
What affects the scope and cost?
The number of decisions, data sources, forecasts and business units matters. So do data quality, required detail, planning horizon, integration needs, documentation and the support your team already has.
I would first separate the work needed to improve the immediate decision from a broader platform or operating change. The initial scope should make its purpose, responsibilities and deliverables clear.
A useful first conversation
Bring one decision that keeps becoming harder than it should, an example of the forecast you currently use and a rough picture of the data behind it. We can discuss the goal and the obstacles before deciding what information or access is actually needed.
Business judgment, analytical clarity and execution
I help make complex work understandable and actionable.
Dr. Robert Urban · Paper Boat Media
My work brings together business strategy, marketing, analytical thinking and the practical challenge of getting people aligned around a decision. I focus on the connection between what the evidence says and what the organization can do with it.
I am a PMP and Certified Scrum Master, with experience in Kaizen, RUP and Waterfall. Those approaches provide different tools for organizing work. The starting point remains a simple, measurable objective, a realistic assessment of the team and a roadmap from the current situation to the intended result.
For forecasting, that means making assumptions explicit, involving the people who understand the operation and defining a process that can survive beyond the first presentation. The best approach is one your organization can use responsibly and maintain.
When the forecasting need connects directly to marketing leadership, resource allocation or commercial planning, fractional CMO support may be part of the wider conversation.
Meet RobFlorida based · National & international work
The market is part of the model.
I am based in DeLand and work primarily across Florida, throughout the United States and internationally. A forecasting conversation should account for the particular customers, locations and operating conditions that shape the business.
A local service company, a professional firm with clients in several states and an industrial supplier with long project cycles need different assumptions. Geography can affect service boundaries, travel time, customer mix, seasonality and the way work is delivered. We identify which factors matter rather than adding location detail simply because it is available.
My markets and geographic growth strategy covers that broader context. Florida work includes DeLand and Volusia County, Orlando, Lake Mary, the Space Coast and other markets across the state. The same decision-focused approach applies when your customers or operations extend much further.
Questions about working together
Forecasting consulting FAQs
We can start with the planning problem you have and determine what kind of support would be useful.
What does a forecasting consultant do?
A forecasting consultant helps a business estimate future activity over a defined period, examine the assumptions and uncertainty, and connect the outlook to decisions. The work can include data review, method selection, scenarios, validation, stakeholder alignment and a repeatable planning process.
How do I choose between forecasting consultants?
Ask how they will define the business decision, assess data quality, establish a baseline, test performance and make the result maintainable. Look for clear explanations of assumptions and limitations, along with a realistic scope and a practical handoff.
Can you help with demand, revenue and staffing forecasts?
Yes. I help connect those questions where they affect the same business decision, while preserving the definitions each function needs. Scope depends on the available evidence, the planning horizon and the analytical or operational support your team already has.
How is forecasting different from predictive analytics?
Forecasting usually estimates future quantities or activity over time. Predictive analytics is broader and can include event probabilities, classifications and risk scores. They can inform each other. The right starting point depends on whether the decision concerns an outlook over time, a particular outcome or both.
We already have dashboards. Do we need forecasting consulting?
Possibly. A dashboard presents information; the forecasting process determines how a future outlook is produced, evaluated and used. If your reporting is strong but planning remains difficult, we can examine what is missing before deciding whether the dashboard itself needs changes.
Do we need AI or machine learning?
Not necessarily. Begin with the business question, a credible baseline and the available data. More complex methods should demonstrate enough practical value to justify their implementation, validation and maintenance requirements.
How much historical data do we need?
There is no universal minimum that fits every forecasting problem. It depends on the frequency, seasonality, number of changes, planning horizon and method. With limited comparable history, we may begin with explicit scenarios, informed judgment and a plan to gather better evidence.
Can you work with our existing analysts, finance team or systems?
Yes. I can provide strategic direction, help define requirements, challenge assumptions and coordinate the work with your team. We agree who builds, reviews, maintains and uses each part of the forecasting process.
Can you guarantee a particular level of forecast accuracy?
No. Performance depends on the data, horizon, business conditions and method. We can define an appropriate evaluation, compare against a baseline, document uncertainty and assess whether the process improves the decisions it is intended to support.
How often should a forecast be updated?
Match the update schedule to the decision, the arrival of meaningful new information and the lead time needed to respond. Some operational questions need frequent reviews; others need a slower rhythm. The refresh process should have clear responsibility and preserve previous versions.
How much does forecasting consulting cost?
Cost depends on the decision scope, data readiness, number of forecasts, integration needs and level of implementation support. We can begin by defining the problem and agreeing on deliverables, responsibilities and the work required.
Do you work outside Florida?
Yes. I am based in DeLand, work primarily across Florida, and serve organizations throughout the United States and internationally. Remote, project-based and retained advisory arrangements can be discussed around your needs.
For readers who want the methodology
Useful forecasting references
The practical recommendations and examples above are written for business decisions. For deeper technical context, Hyndman and Athanasopoulos’ Forecasting: Principles and Practice provides the following explanations.
- Simple forecasting methods and benchmarks · Establishing a meaningful reference.
- Time-series cross-validation · Evaluating forecasts across historical decision points.
- Evaluating forecast accuracy · Understanding measures and their limitations.
- Prediction intervals · Communicating model-based uncertainty.
- Scenario forecasting · Exploring conditional alternatives.